Clinical standard for valve area after common atrioventricular valve plasty for a single ventricle

Abstract

OBJECTIVES

To determine a clinical standard for post-repair common atrioventricular valve orifice area based on mid- to long-term valve function in patients with a functional single ventricle.

METHODS

The medical records of 19 single-ventricle patients who underwent common atrioventricular valve plasty from July 1988 to January 2013 were retrospectively reviewed. Bivalvation valvuloplasty was performed in 7 patients with relatively hypoplastic leaflets. The relationship between the orifice area of the repaired common atrioventricular valve measured intraoperatively and valve function and ventricular volume in the early postoperative period (median, 9.5 months) and at mid- to long-term follow-up (median, 4 years) were analysed.

RESULTS

Post-repair valve area was significantly positively correlated with valve regurgitation severity in the early postoperative period (P = 0.001, r = 0.69) and at mid- to long-term follow-up (P = 0.02, r = 0.57). Patients who did not undergo bivalvation had favourable valve function at mid- to long-term follow-up and in the early postoperative period when the post-repair valve area was 96–136% of the normal mitral valve area. Bivalvation patients had significantly more valve regurgitation in the early postoperative period than patients without bivalvation, despite equivalent repaired valve area (P = 0.02).

CONCLUSIONS

The post-repair orifice area of the common atrioventricular valve is significantly related to postoperative valve function. The clinical standard of post-repair valve orifice area might be 96–136% of the normal mitral valve area in patients undergoing repair without bivalvation. Patients undergoing bivalvation require greater reduction to obtain favourable mid- to long-term valve function.

INTRODUCTION

Significant atrioventricular valve regurgitation increases mortality in patients with single-ventricle physiology [1–4]. Among several types of atrioventricular valves, common atrioventricular valve (CAVV) is associated with the most difficult-to-regulate regurgitation (CAVVR); the presence of CAVVR is an independent predictor of adverse outcomes [4]. Mechanisms of CAVVR in single-ventricle physiology are complex and multifactorial [5, 6]. The main causes of CAVVR are congenital abnormalities of the leaflet and of subvalvular structures [3, 7], although annular dilatation can also contribute to regurgitation [1, 8]. The annular diameter of the CAVV in patients with a single ventricle who had not undergone atrioventricular valve repair is reportedly closely related to the severity of CAVVR and ventricular volume, indicating that the annular diameter of the CAVV can predict prognosis [9]. Although the repaired CAVV orifice area might help determine surgical outcomes, this is as yet unclarified; the clinical standard for repaired CAVV orifice area in patients with a functional single ventricle remains undetermined.

Various procedures are performed for CAVV plasty (CAVVP) [10]. Among these, bivalvation is unique, because it produces a double orifice and is often applied to abnormal valve morphology with extensively reduced leaflet coaptation [11]. The impact of surgical reduction of CAVV orifice area could be different in patients who undergo bivalvation than in those repaired without bivalvation.

The present study investigated the clinical standard for repaired CAVV orifice area based on mid- to long-term valve function in patients with a functional single ventricle. We also assessed outcomes of bivalvation versus repair without bivalvation, with equivalent repaired valve orifice area.

MATERIALS AND METHODS

Patients

This retrospective multicentre study was approved by the institutional review board of Osaka University Hospital (approval number 15241-1). The application of the Congenital Osaka Cardiovascular Research Group, which consists of Osaka University Hospital, Osaka Medical Center and Research Institute for Maternal and Child Health and Osaka City General Hospital, was approved by each hospital. From July 1988 to January 2013, 23 patients with a functional single ventricle and CAVVR underwent CAVVP at the above 3 hospitals. Four patients were excluded because detailed operative information was unavailable. The study included 19 patients (9 males; median age at initial CAVVP, 1.2 years). Patient characteristics are shown in Table 1. The preoperative degree of CAVVR was mild to moderate in 8 patients, moderate in 7 and severe in 4.

Surgical procedure

CAVVP data are shown in Table 2. Three patients underwent CAVVP before Glenn operation. Of the 13 patients who underwent CAVVP concomitantly with bidirectional Glenn operation, antegrade pulmonary blood flow persisted in 5. Three patients underwent CAVVP with non-fenestrated Fontan operation. At Fontan completion, all patients underwent total cavopulmonary connection with an extracardiac conduit. All procedures were performed through a median sternotomy, under cardiopulmonary bypass with mild-to-moderate hypothermia and cardiac arrest using aortic cross-clamping.

Bivalvation was performed in 7 patients with annular dilatation combined with a relatively hypoplastic leaflet; the estimated primary aetiologies are listed in Table 3. The free edges of both bridging leaflets were simply sutured in 4 patients, while a bridging strip made with an expanded polytetrafluoroethylene graft was attached to reduce the anteroposterior width of the bridging leaflet to approximately 80% of the original distance in 3 patients. Additional surgical techniques for CAVVP were concomitantly performed in all 7 patients who underwent bivalvation. These techniques included commissural annuloplasty in 3 patients, paracommissural repair (partial closure of a commissure with decreased coaptation) in 2, DeVega annuloplasty in 1 and DeVega annuloplasty with concomitant commissure annuloplasty in 1. Remaining 12 patients underwent repair without bivalvation.

CAVV orifice diameter was measured intraoperatively with a bougie; the area was calculated as a percentage of normal mitral valve (MV) area, assuming that the orifice was a complete circle. For patients who underwent bivalvation, the CAVV orifice area was defined as the sum of the areas of the divided CAVV orifices. Normal MV area was calculated from the normal MV diameter, assuming that the valve was a complete circle. For intraoperative measurement of valve orifice diameter, normal MV diameter was estimated as 24.3 × BSA^0.44 [12]. For echocardiographic measurements, normal MV diameter was estimated as 25.72 × BSA^0.5022 [13].

The techniques used for CAVVP were selected based on preoperative echocardiography and intraoperative saline injection testing. Bivalvation was mainly selected when the valve had relatively hypoplastic leaflets, and it seemed difficult to obtain enough coaptation with repair methods other than bivalvation. After confirmation of valve competence with the intraoperative water test, the valve orifice area was calculated as described above. To avoid valve stenosis, we attempted to preserve an orifice area greater than 100% of normal MV area. CAVVP was performed using a suitable technique according to valve structure without any established target area for the repaired valve orifice other than this lower limit.

Study methods

Routine cardiac catheterization was conducted before Glenn operation, before Fontan operation and 1 year after Fontan operation. The following data were collected from medical records, which included echocardiography and catheter examination reports from the preoperative period, early postoperative period and mid- to long-term follow-up period: (i) overall outcomes (actuarial survival rate and probability of Fontan completion) and (ii) serial changes in CAVV diameter, CAVV function, atrial pressure, ventricular pressure and ventricular volume.

The pressure gradient across the CAVV was measured, with catheter examination at a median of 1.5 months preoperatively, 9.5 months postoperatively (early postoperative period) and 4 years postoperatively (mid- to long-term follow-up). To eliminate differences in ventricular volume load according to surgical stage, systemic ventricular end-diastolic volume index (SVEDVI) was evaluated only in patients who underwent CAVVP at the Glenn stage. SVEDVI was measured with catheter examination at a median of 100 days before Glenn operation, 324 days after Glenn operation (early postoperative period) and 413 days after Fontan operation (mid- to long-term follow-up). In 2 patients who underwent repeat CAVVP for ongoing CAVVR at the Fontan stage, data from catheter examination at a mean of 83 days before Fontan operation were evaluated.

Echocardiographic examination of CAVV was performed concurrently with catheter examination and at the latest follow-up (median, 8 years after CAVVP). Paediatric cardiologists evaluated CAVV function and CAVV orifice diameter with echocardiography. CAVV orifice diameter was measured using the 4-chamber view. The CAVV orifice diameter after bivalvation was calculated as the square root of the sum of the square of each valve diameter. The degree of CAVVR was evaluated with Doppler colour-flow mapping and described as follows: none = 0, trivial = 1, mild = 2, moderate = 3 and severe = 4. These variables were compared with the post-repair CAVV orifice area. No-to-mild CAVVR was considered acceptable valve function; greater than mild-to-moderate CAVVR was considered uncontrolled regurgitation.

Statistical analysis

Data are expressed as median (range). Actuarial survival was estimated with the Kaplan–Meier method. Associations between the CAVV orifice area (expressed as the percentage of normal MV area) and other variables were evaluated with the paired t-test and Spearman’s rank correlation coefficient. The Wilcoxon test was used for data without normal distribution. Comparison of 3 groups of repeated measurements was performed using repeated-measures analysis of variance with Bonferroni’s method. Regression lines were compared with analysis of covariance. All data were analysed using JMP Pro 12 (SAS Institute, Cary, NC, USA). Differences were considered statistically significant at P-values <0.05.

RESULTS

Overall outcomes

CAVV orifice diameter was reduced with surgical plasty from 26 mm (20–44 mm) to 19 mm (10–24 mm) (P < 0.001), and CAVV orifice area was reduced with surgical plasty from 250% (200–402%) of normal MV area to 137% (96–190%) (P < 0.001). All patients completed the follow-up; the median follow-up period was 5.7 years (4 days–27 years). The actuarial survival rate after initial CAVVP was 63.2% at 5 years and 57.8% at 10–20 years. There were 6 in-hospital deaths (31%). All in-hospital mortality patients had right isomerism. Among these, 3 patients had persistent uncontrollable CAVVR. Two of these 3 patients died of heart failure and the third died of pulmonary hypertension crisis. One patient needed extracorporeal membrane oxygenation (ECMO) after surgery and died with progression of systemic oedema. One patient died of pulmonary bleeding and another died of sepsis despite recovery of CAVV function. There were 2 late deaths. Both these patients had right isomerism and underwent CAVVP at Glenn stage, followed by Fontan operation. One of these patients died at 1 year and the other at 3 years after Fontan operation due to unimproved heart failure with moderate CAVVR. Among all 8 mortality cases, uncontrollable CAVVR persisted in 5 patients with a post-repair CAVV orifice area of 139–190% of normal MV area. Two patients had resolution of CAVVR postoperatively; their respective post-repair CAVV orifice areas were 125% and 95% of normal MV area. Information was unavailable regarding the degree of postoperative CAVVR in the patient who needed postoperative ECMO support. The CAVV orifice diameter measured with echocardiography in 13 hospital survivors was 25.3 mm (21.0–32.9 mm) preoperatively, 23.5 mm (18.5–32.6 mm) at the early postoperative period and 28.0 mm (22.5–46.0 mm) at mid-to-long term follow-up. The CAVV orifice diameter was significantly decreased post-repair when expressed as a percentage of normal MV diameter (P = 0.001) (Table 4).

Four patients underwent reoperation for persistent CAVVR. One patient whose CAVV orifice diameter was reduced from 20 mm to 16 mm (CAVV orifice area reduced from 300% to 123% of normal MV area) with paracommissural repair experienced ruptured chordae tendineae 10 months after CAVVP and underwent CAVV replacement 4 years after CAVVP. Repeat CAVVP was performed at a mean interval of 2.3 years after initial CAVVP in 1 patient whose CAVV orifice diameter was reduced from 28 mm to 20 mm (CAVV orifice area reduced from 304% to 155% of normal MV area) with DeVega annuloplasty and 2 patients whose respective CAVV orifice diameters were reduced from 33 mm and 31 mm to 23 mm and 21 mm (respective CAVV orifice areas reduced from 358% and 244% to 138% and 130% of normal MV area) with bivalvation. Reoperation resulted in hospital death in 1 patient and late death in 1 patient. Another patient who underwent bivalvation and had a post-repair CAVV orifice area of 130% of normal MV area showed improvement of CAVV function after additional annuloplasty with repeat CAVVP. Fontan operation was completed in 12 of the 19 patients (63%). Individual patient outcomes after CAVVP are summarized in Fig. 1.

Degree of common atrioventricular valve regurgitation

The degree of CAVVR was evaluated via echocardiography in 19 patients preoperatively, 18 in the early postoperative period and 12 hospital survivors at mid- to long-term follow-up. One in-hospital mortality patient was excluded from postoperative evaluation because she needed ECMO support immediately after surgery. One hospital survivor was excluded from evaluation at mid- to long-term follow-up because of ruptured chordae tendineae 10 months after CAVVP. A tendency of improved valve function was seen in serial changes in the degree of CAVVR (P = 0.056) (Fig. 2A). There were significant positive correlations between post-repair CAVV orifice area and severity of CAVVR in the early postoperative period (P = 0.001, r = 0.69) and at mid- to long-term follow-up (P = 0.02, r = 0.57) (Fig. 2B and C).

Systemic ventricular end-diastolic volume

SVEDVI was evaluated with cardiac catheterization in 11 hospital survivors who underwent CAVVP at the Glenn stage. Preoperative data from 1 patient and mid- to long-term follow-up data in 1 patient were unavailable and are not shown. Mid- to long-term follow-up data in the patient with ruptured chordae tendineae are not shown. Analysis of variance showed no significant difference in serial change of SVEDVI (P = 0.55) (Fig. 3A). Three patients with uncontrolled CAVVR showed progressive dilatation of the systemic ventricle after CAVVP. SVEDVI in the early postoperative period was weakly correlated with the size of the repaired CAVV orifice (P = 0.07, r = 0.56) (Fig. 3B). SVEDVI was significantly positively correlated with post-repair CAVV orifice area at mid- to long-term follow-up (P = 0.02, r = 0.80) (Fig. 3C).

Pressure gradient across the common atrioventricular valve

The pressure gradient across the CAVV was evaluated with cardiac catheterization in 13 hospital survivors. CAVV stenosis was clinically insignificant in all patients during follow-up (mean 0.5 ± 0.9 mmHg at the early postoperative period, mean 1.5 ± 0.9 mmHg at long-term follow-up, P = 0.06). In 5 hospital mortality cases, CAVV function was evaluated with echocardiography at a median of 34 days postoperatively; none of these patients showed significant CAVV stenosis. CAVV function could not be precisely evaluated in 1 hospital mortality patient who needed ECMO support immediately after surgery.

Comparison of common atrioventricular valve function after bivalvation versus repair without bivalvation

There was no significant difference in the baseline or post-repair CAVV orifice area between patients who underwent bivalvation versus those who did not. The degree of CAVVR in the postoperative period was significantly greater in patients who underwent bivalvation (P < 0.001) (Table 3).

The degree of CAVVR in the early postoperative period was compared in patients who underwent bivalvation versus those who did not. There was no significant correlation between post-repair CAVV orifice area and surgical procedure (P = 0.28). Patients who underwent bivalvation had significantly greater CAVVR in the early postoperative period compared with patients who underwent repair without bivalvation, despite equivalent post-repair CAVV orifice area (P = 0.02) (Fig. 4A).

Among patients who underwent bivalvation, only 1 patient, who had a post-repair CAVV orifice area of 96% of normal MV area, had acceptable postoperative CAVV function (Fig. 4B). Among the patients who underwent repair without bivalvation, the post-repair CAVV orifice area was significantly larger in 3 patients who had uncontrolled CAVVR (median, 147 ± 9% of normal MV area) compared with the other 8 patients who had acceptable valve function in the postoperative period (median, 125 ± 12% of normal MV area) (P = 0.017) (Fig. 4B). Among the patients who underwent repair without bivalvation, all patients with a post-repair CAVV of 96–136% of normal MV area had acceptable CAVV function during follow-up (Fig. 4B).

DISCUSSION

The major findings of this study were: (i) the intraoperative-repaired CAVV orifice area was significantly related to postoperative CAVV function; (ii) bivalvation repair was associated with significantly greater CAVVR in the early postoperative period than repair without bivalvation, despite equivalent post-repair CAVV orifice area; (iii) among the patients who underwent repair without bivalvation, all patients with post-repair CAVV orifice area of 96–136% of normal MV area had favourable valve function at mid- to long-term follow-up as well as in the early postoperative period, while postoperative CAVV function was suboptimal in all patients that underwent bivalvation, except for 1 with a post-repair CAVV orifice area of 96% of normal MV area.

MV annulus dilatation reportedly predicts functional MV regurgitation in adults [14, 15]. In congenital heart disease, Sano et al. analysed the influence of CAVV annular diameter on CAVVR and ventricular volume in 24 patients with functional single ventricles who had not undergone atrioventricular valve repair [9]. They reported that patients with CAVV annular diameter over 130% of normal MV diameter tended to have deteriorated ventricular function, relatively severe CAVVR, and a poor prognosis, while patients with annular diameter under 115% of normal MV diameter had preserved ventricular function with mild or less CAVVR [9]. The mechanism of CAVVR in single-ventricle physiology is multifactorial, and valve size is not the sole determinant factor for CAVV function [3, 5–7]. However, the present study indicates that post-repair CAVV orifice area is an important predictor of valve function at mid-to-long-term follow-up, as demonstrated by the significant positive correlation between postoperative degree of CAVVR and post-repair CAVV orifice area which was measured after confirmation of valve competence.

Bivalvation for CAVVR is advocated to reduce the central superior–inferior annular dimension to improve coaptation of the bridging leaflets [3, 11]. This procedure is often indicated in patients with annular dilatation combined with extensively reduced leaflet coaptation [11]. In the present study, the post-repair CAVV orifice area was measured intraoperatively with a bougie, assuming that each orifice was a complete circle. This technique possibly overestimated the post-repair CAVV orifice area as the orifice can become flexible under cardiac arrest, allowing a larger bougie to pass. Moreover, the sum of each orifice area evaluated independently might be larger than the true value. Even though the post-repair CAVV orifice area was potentially overestimated in patients who underwent bivalvation, these patients had significantly greater CAVVR in the early postoperative period compared with patients who underwent repair without bivalvation, despite equivalent post-repair CAVV orifice area and primary valve abnormalities. Bivalvation was mainly selected when the valve had relatively hypoplastic leaflets, and it was difficult to obtain enough coaptation with isolated valve plasty. Therefore, the poorer valve function after bivalvation might be related to the unfavourable form of the valve itself rather than the surgical technique.

The optimal CAVV orifice area in single-ventricle physiology might be determined by the volume of blood flow passing through the CAVV orifice. Therefore, the optimal CAVV orifice area might be a sum of 100% of normal MV area for the systemic blood flow plus the area for pulmonary blood flow originating from the systemic ventricle. Based on this theory, in patients after Glenn operation without additional pulmonary blood flow and Fontan operation, the optimal CAVV orifice area is 100% of the normal MV area. Otherwise, it is greater than 100% of the normal MV area. While the valve stenosis may develop when the post-repair CAVV orifice area is less than this optimal size, no patients in the present study showed significant valve stenosis after CAVVP with post-repair CAVV orifice area of 96–190% of normal MV area.

Limitations

Our study has several limitations. First, this was a retrospective study with a small sample size. We did not have enough patients to analyse the impact of some of the parameters studied, including morphologic diagnosis, effect of antegrade pulmonary artery flow or collateral systemic-to-pulmonary flow and timing of CAVVP. Second, the cohort consisted of heterogeneous patients with different cardiac morphologic types who underwent various surgical procedures at different stages. Third, details of repaired valve function were not revealed, except for the pressure gradient across the CAVV, grade of regurgitation and ventricular volume. Fourth, although the shape of the CAVV associated with atrial isomerism is typically elliptical rather than circular, the orifice was assumed to be a complete circle in calculation of the CAVV orifice area, which might result in the overestimation of the valve area. Fifth, the lower size limitation of the optimal CAVV orifice area could not be identified because no patients showed valve stenosis.

CONCLUSION

In conclusion, post-repair CAVV orifice area measured intraoperatively was significantly related to postoperative valve function. Intraoperative evaluation of the post-repair CAVV orifice area is important to predict outcomes in patients with a functional single ventricle and CAVVR. The clinical standard of post-repair valve orifice area was 96–136% of normal MV area in patients who underwent repair without bivalvation. More drastic reduction was needed in patients who underwent bivalvation, and reduction to the size corresponding with cardiac preload might be insufficient to obtain favourable mid- to long-term valve function.

Conflict of interest: none declared.

REFERENCES